Composite material production process parameter optimization method based on digital twinning

Abstract

This invention relates to the field of simulation technology, specifically to a method for optimizing composite material production process parameters based on digital twins. This method constructs a digital twin space encompassing equipment status, material history, and environmental parameters. A spinning and drawing simulation model is established using the Lagrange-Euler method and multibody dynamics, and rheological parameters are dynamically calibrated using measured data to generate a dynamic mechanism. CFD simulation and virtual sensor algorithms are used to fuse physical measurement data to construct a global environmental field map. A physically constrained variational autoencoder is used to extract cross-scale correlation features between the dynamic mechanism and the environmental field. Within a probabilistic latent space with Fisher information metric, unsteady-state modes deviating from physical criteria are identified, and the mechanisms of fiber breakage and fuzz formation are predicted in reverse, generating latent quality risks. Finally, a genetic algorithm is used to pre-fracture and optimize compensation paths within the twin space, generating adjustment commands that are energy-efficient and meet quality indicators.

Images